24 October MEDCL JOURNL 15 Ppers nd Originls Nturl History nd Tretment Wilms's Tumour : n nlysis 335 Cses Occurring in Englnd nd Wles 12- E. M. LEDLE,* M.B., B.S., D.M.R.; L. S. MYNORS,* M.B., B.S.; G. J. DRPERjt M..; P. D. GORBCH4t B.. British Medicl journl,, 4, 15-2 Summry: Hospitl records nd other dt relting to Wilms's tumour were nlysed to elucidte both the nturl history the disese nd the effects tretment, with specil reference to ctinomycin D. The ge mximum incidence ws bout 18 months; left-sided tumours were significntly more common thn right-sided ones. Prognosis ws relted to stge the disese t the initil opertion nd to the occurrence metstsis. The threeyer survivl rte for cses hving nephrectomy ws 35%; no child who did not hve nephrectomy survived. Recurrence the tumour ws observed in two-thirds the cses, lmost lwys within two yers the initil tretment, irrespective the child's ge. The three-yer survivl rte for this group ws 11%. The effects rdiotherpy nd chemotherpy re considered in detil. Very little improvement in survivl rte could be scribed to ctinomycin D. The resons for this nd for the vritions found in erlier reports on selected cses re possibly the ddition other components tretment nd differences in drug regimens. The findings suggest the need for controlled clinicl trils. ntroduction For some yers the Oxford Survey Childhood Cncers hs been collecting dt on ll cses children certified s dying from cncer nd lso cses notified to registries in Englnd nd Wles who hve survived. These dt hve mde it possible to study individul diseses, first from the point view their nturl history, nd, secondly, in order to ssess the effectiveness different tretments. The present study, which is probbly unique in tht every cse, including untreted nd coroners' cses, is included, is bsed on 335 cses Wilms's tumour, nd is thought to comprise ll the children in Englnd nd Wles under the ge 1 yers who were dignosed during the five-yer period 12-. No cse without histologicl confirmtion hs been included. The informtion required for the ded children ws obtined from the hospitl notes, while for most the surviving children the generl prctitioners interviewed the mothers nd loned to us the relevnt hospitl letters. Consultnts were sked to supply the necessry detils for the reminder. On checking the records ginst the deth certifictes the inccurcy the ltter ws found to be bout 8 %. Nineteen cses certified s Wilms's tumour hd to be discrded-two-thirds becuse the necropsy findings did not confirm the deth certificte dignosis, nd one-third on ccount copying errors owing to the similrity the words "neuroblstom" nd "nephroblstom." Nturl History The incidence nephroblstom ws bout 13'Y,, ll children under 15 yers suffering from cncers other thn * Reserch ssistnt. Lecturer in Medicl Sttistics. t Sttisticl Reserch ssistnt. Deprtment Socil Medicine, 8 Keble Rod, Oxford University, Oxford OX 3QN. leukemi during this period, figure which is similr to those published from Gret Ormond Street (Riches, 14), Mnchester (Person nd Stewrd, 1), nd the Cliforni Tumour Registry (Pltt nd Linden, 14). There were more boys thn girls (184 boys, 151 girls; rtio 1-2: 1) (Fig. 1) nd more left-sided thn right-sided tumours (185 left, 13 right; rtio 1-3:1; lso nine bilterl cses). significnt excess left-sided tumours ws found t the 2% level nd ws confined to cses without extrrenl spred. Hlf the tumours presented before the ge 3 yers, the pek incidence being t 18 months. The totl smple could be divided into three lmost equl ge groups, 11 under 2 yers t dignosis, 115 ged 2 or 3, nd 1 over 4 yers (no record one). These figures re similr to those in other published series (Collins, 155; Young nd Willims, 1). The Registrr Generl's figures for deths from Wilms's tumour in children under 1 for the period 14-8 showed only slight prepondernce boys, 87 mle deths s compred with 85 femle deths. This is lower sex rtio thn tht found in our series; the period covered by the present study is typicl in this respect. The modl ge cses rising spontneously is consistent with Stewrt nd Knele's () finding tht for tumours which were probbly cused by obstetric rdiogrphy, nd therefore initited during the third trimester (rdiogenic cses), the commonest ltent period or intervl from initition to the onset symptoms ws 12 to 3 months. Congenitl tumours occurred in 1 children, six whom hve survived for three yers or more without developing signs further disese. ll 1 underwent nephrectomy, nd the four deths occurred within the first month with no evidence spred beyond the primry site. The bsence metstses in ny this group, together with the unusully high survivl rte, rises the possibility tht the tumours were fetl mesenchyml hmrtoms (Wigger, 1). The nine bilterl cses hve been clssified in wy modified from tht suggested by Scott (155): found 7 - - s5 ^o 4 M 3- E 2 Z nor 1* 2 3 4 5 ge t d ignos s ( ye rs ) Do OX--% 7 8 K) FG..-Wilms's tumour in Englnd nd Wles 12-: sex relted to ge t dignosis.
1 24 October ctinomycin D in Wilms's Tumour-Ledlie et l. BlUTSH MEDCL JOURNL primrily t necropsy, 2 cses; found t first opertion, 5 cses (one these hs survived over four yers without recurrence, the other four hve died); nd unilterl t first opertion with second tumour found lter, 2 cses. The 241 cses for which the sibship ws known included nine boy twins nd two girl twins. There ws one pir identicl mle twins, whom one died nd the other hs survived without recurrence for t lest four yers fter his tretment t the ge 1 yer. The incidence congenitl defects ws similr to tht lredy described by Brber nd Spiers (14). The defects found in the present series re summrized in Tble. TBLE.-Congenitl Defects ssocited with Wilms's Tumour Mle Femle Totl Genitourinry-hypospdis, duplex kidney, nd polycystic disese 4 3 7 Hemihypertrophy 2 3 5 niridi..1 1 Other.4 3 7 Totl.1 1 2 niridi occurred in one child only (cf. n incidence 1:73 reported by Miller (18) for children with Wilms's tumours). She ws mentlly retrded child with bilterl niridi nd lens opcities. The nephroblstom, dignosed t the ge 1 months, ws treted by nephrectomy, but the ptient died four months lter. History nd Presenting Symptoms The durtion symptoms ws known in 27 the ftl cses nd in 83 the survivors. One-third the whole series were dmitted within week nd two-thirds within month. No correltion between the length history nd the extent the disese ws found. mong the 8 cses (one-third the whole) with symptoms lsting month or more before tretment ws instituted dely up to four weeks occurred in 18 cses fter doctor hd been consulted before lprotomy ws done. Eleven these presented with hemturi nd were kept under observtion s probbly hving cute nephritis; one ws found to hve duplex kidney nd not investigted further; nd the other six were given provisionl dignoses "enlrged spleen" (three cses), convulsions, obscure bdominl pin, nd umbilicl herni. The remining 8 cs E 11, 8 7.- 4.- 3.J x Recur * Recur rence-live t ltest follow-up, No re( currence-live t ltest follow-up s,'. c rence - ded / D / ~ C 8 o o o p ) CooP o o~ p ' Co, - end 2-1er -- --- X--------x-X------------- -zb- - X.. riod ris g x* x x~xxxwxx ox, x x x XX ~~~~~~~~~~~~x 2 3 4 5 7 8 1 ge t dignosis (yers) FG. 2.-Recurrence in Wilms's tumour. For the explntion "periods risk" see text. x cses with long history hd incresing bdominl distension, pin, mlise, or vomiting, nd the history ten ended with sudden crisis pin, distension, or hemturi which brought them quickly to hospitl. There seems therefore little reson to suppose tht more could be done to hsten the dmission to hospitl children with nephroblstoms. n most cses n bdominl mss ws plpble on dmission nd tretment ws promptly instituted. bdominl pin ws present in one-third the cses. Hemturi nd generl symptoms (pyrexi, lssitude, norexi) ech occurred in bout one-qurter the cses. Eighteen children gve history bdominl symptoms following n injury, eight were dmitted s cute bdominl emergencies, nd five hd no symptoms t ll, the tumour being discovered by the doctor exmining the child for n unconnected symptom. Recurrence Tumour The term "recurrence" will be tken to men reppernce neoplstic tissue either loclly or outside the bdomen fter removl the primry tumour. Two hundred nd thirty-two children thought to be tumour-free fter nephrectomy hve been followed up until deth or for survivl period usully exceeding three yers. The intervl to recurrence or lst follow-up is plotted ginst ge t dignosis for these children (Fig 2). recurrence ws observed in 15 cses (7%), whom 141 hve died (three them fter surviving three yers). The three-yer survivl rte for these cses ws 12%. No recurrence hs been observed in 7 cses, ll whom hve survived. The length the so-clled "period risk"-tht is, the period beyond which recurrence the disese is unlikelyhs been discussed by mny uthors, including Collins et l. (15) nd Pltt nd Linden (14). Collins suggested tht the period risk ws the ge the child plus nine months, on the grounds tht the originl tumour probbly strted close to the time conception nd ny recurrence might be expected to hve the sme growth rte. n Fig. 2 the period risk postulted by this rule is mrked f by the digonl line. f, however, it is postulted tht the period risk is usully two yers, irrespective the ge the child, the period is delimited by the horizontl line. n four cses the intervl to recurrence exceeded the period given by the ltter rule; these, three occurred within 28 months dignosis. Two the four lso exceeded the period given by Collins, but since both hd bilterl tumouxs it is probble tht these "recurrences" were in fct new tumours. One child, fter nephrec- 4) F- 8 7 " 4 w ': 11 1 4 c _~~~ l? 1 lt L. :11, : 4 :, 11 2 3 4 qe t dignosis (yers) Q 5 7 8 FG. 3.-Survivl fter recurrence in Wilms's tumour. The verticl lines show the length time between the occurrence metstses, indicted by *, nd the time ltest follow-up t which the child ws live, indicted by
24 October ctinomycin D in Wilms's Tumour-Ledlie et l. BRTSH MEDCL JOURNL Lungs: Bilterl nd multiple Left midzone. Solitry Bilterl. 2 Deposits? Site. Solitry Right side. 3 deposits Bilterl nd multiple R.U.L. Solitry Right side. Multiple Right pex. Solitry R.L.L. Solitry L.U.L. nt. medistinum L.L.L. Solitry bdomen: Opposite kidney Opposite kidney Liver Lungs: R.U.L. Solitry Bilterl nd multiple Bilterl nd multiple TBLE.-Wilms's Tumour-Metstses in Survivors Tretment letstses ppering fter First Tretment ctinomycin Vincristine. XRT 2 Lobectomiess. ctinomycin Lobectomy. C:yclophosphmide Excision. Vinc-ristine. ctinomycin XRT only XRT only Excision. XRT. ctinomycin Lobectomy. ctinomycin Wedge resection. ctinomycin XRT. ctinomycin B Lobectomy C Excision. ctinomycin XRT. ctinomycin L. heptic lobectomy B Metstses Present t First Tretment Excision. XRT XRT. ctinomycin. Cyclophosphmide XRT. ctinomycin *Stge R.U.L. = Right upper lobe. L.L.L. =Left lower lobe, etc. XRT = Rdiotherpy. *See "Prognostic Fctors" for definition. C ge t ge t Follow-up fter Dignosis Recurrence Recurrence (Months) (Months) (Months) 23 25 42 48 1 2 7 42 55 88 2 13 41 28 1 1 72 33 32 4 57 8 8 7 7 1 5 1 3 38 5 48 1 1 72 34 3 35 43 2 7 43 37 73 37 25 38 24 4 28 51 41 3 tomy t 13 months, ws found to hve tumour in her remining kidney 25 months subsequently. The second child hd right nephrectomy nd left heminephrectomy for bilteirl tumours when 23 months old. tumour developed 53 months lter in the remining kidney tissue. The pttern recurrence times strongly suggests tht beyond two yers the chnce recurrence is very smll, nd tht the two-yer rule is more useful from prcticl point view thn tht suggested by Collins. One my in fct question the ppliction Collins's rule to tumour where the recurrence tkes the form metstsis, since the ssumptions similr growth rtes nd known time initition do not pply to this sitution. Only 2 the 141 ftl cses lived more thn two yers from the time recurrence. This suggests tht the 15 who hve survived longer thn this hve been successfully treted for their recurrences nd re cured. Detils these 15 nd lso three surviving children who presented with metstses re given in Tble. The intervl to recurrence nd length subsequent follow-up re shown in Fig. 3. Sites Metstses nlysis the site metstsis dignosed before deth or found t necropsy hs been ttempted, but is only pproximte, for in some children neither the sites nor their full extent were known. () Pulmonry Metstses.-n totl 15 children lung deposits only were detected. further 3 cses in which the intrthorcic metstses were prt widespred disese re considered under (c). The most striking feture the lung 5-41 or v 3 o. M 3-yer survivors M Deths B c D E None Lungs bdomen Multiple Unknown metstses ws the rise in incidence nd proportionte survivl with incresing ge the children. The reltive incidence nd three-yer survivl children with metstses in the lung nd elsewhere is shown in Fig. 4. (b) bdominl recurrence (s distinct from extrcpsulr spred the tumour) occurred in 53 cses nd ws chrcterized by being found lmost eqully in ll ge groups nd crrying mortlity % t ll ges. (c) Metstses in multiple sites were found in 4 cses, ll which were ftl. Vrying nd unpredictble combintions sites occurred, though in more thn hlf the cses the sites involved were lungs nd intr-bdominl structures, including the liver. n two cses bone metstsis (skull nd zygom) ws the first sign reppernce ctive disese, but in these nd in ll other cses with bony deposits other orgns were subsequently ffected. Bones involved were ribs, pelvis, spine, skull, nd zygom. Two children born with hydrocephlus died within the first yer nd were found to hve Wilms's tumours with intrcrnil deposits. Prognostic Fctors Of the totl 335 cses, (32%) were live three yers fter dignosis. (The four surviving cses with periods follow-up between 3 nd 35 months hve been counted s three-yer survivors.) The survivl curve for intervls up to three yers-tht is, the proportion cses still live t ech time-is shown in Fig. 5. Most the deths occurred within two yers dignosis. There is no difference between the sexes s regrds survivl. The extent bdominl spred ws clssified ccording to 1 8 _ 2 4 '~ 1 2-1 ge (yers) 2-3 FG. 4.-Frequency metstses in Wilms's tumour relted to ge t dignosis. ll qes, Months fter dignosis FG. 5.-Survivl curve for 335 cses Wilms's tumour dignosed in Englnd nd Wles 12-.
18 24 October ctinomycin D in Wilms's Tumour-Ledlie et l. BRMnTs MEDCL JOURNL the following scheme: stge, tumour encpsulted with no extension the tumour outside the cpsule, though the renl prenchym my be involved; stge B, dherence to surrounding structures, but no invsion; nd stge C, intrbdominl spred outside the kidney, or rupture tumour during removl. These three stges re modified from stges,, nd Grci et l. (13), who found tht they were considerble prognostic significnce. The three-yer survivl rtes for ech stge, ech gegroup, nd lso ccording to the presence or bsence metstses t the time opertion nd lter re shown in Tble. Ech these fctors is strongly relted to the probbility survivl. bout two-thirds stge cses, but very few stge C cses survive; older children hve worse prognosis thn younger ones; only smll proportion cses with metstses survive, nd this is true whether the metstses were discovered t the time opertion or lter. These three fctors, however, re closely interrelted, nd more detiled exmintion the dt is needed to elucidte their seprte effects. The proportion cses with metstses increses with the extent bdominl spred (Tble V). Within the stge group ge is relted to the occurrence metstses nd to survivl. n this group the proportion cses with metstses is 2 % for those under 2 yers, 4 % for those ged 2 nd 3 yers, nd 74 /, for those ged 4 yers or more. The overll survivl rte for these cses is 2%, wheres the rte for stge cses without metstses is 1,. This ccounts for the reltionship ge to survivl in this group shown in Tble V, which lso shows tht the older children re more likely to be stge B nd C nd hence to hve worse prognosis. The chnce survivl therefore depends on the stge the tumour t dignosis, nd on whether or not metstses occur. The reltionship survivl to ge rises from the fct tht older children re more likely to hve tumours t more dvnced stge nd re more likely to show metsttic involvement even with loclized tumours. nitil Tretment mong the 34 children subjected to nephrectomy (% the cses) the overll three-yer survivl rte ws 35 %: no ptient who did not hve nephrectomy survived. ssessment the effect dding rdiotherpy nd chemotherpy to nephrectomy is difficult, since the groups cses receiving different tretment combintions re not comprble. Furthermore, fctors known clinicl significnce, such s histologicl grde nd size tumour, which were not known, TBLE.-Three-yer Survivl Rtes for Wilms's Tumour by Stge, ge t Dignosis, nd Occurrence Metstses. Cses where the necessry dt re not vilble re omitted from this tble ll cses.. 335 r.. Stge B. ro-i ge t dignosis in yers { 42-3 t4- Metstses: Present t opertion.. Found lter. ll cses with metstses None found.. N1 No. 3-Yer NCo Cses live 3 Yers t Rte Survivl (%).. 121 78 3 125 8.. 11 4 115 32 1 2 Stge 35 158 28 83 Stge B 32 3 18 21 7t 1~~-----1-- Distnt metstses present: t opertion.. 5 4 3 Lter. 4 38 2 77 ll cses with metstses 53 44 24 8C None found.. 4 53 5 7 Dt not vilble.. 4 3 1 3 ll cses.. 121 F 1 3 1 1 *ncludes cses where the time detection ws not recorded. 4 3 41 28 2 11 1 2t *ncludes cses where time detection ws not recorded. tthe seven children without metstses who died were those who did not hve nephrectomy. f these re excluded the survivl rte for cses with no metstses is 1%. TBLE V.-Frequency Metstses in Wilms's Tunmour Detected t Opertion or Lter, Relted to Stge Stge C No., No. ",, - Cses Cses NoCs " Cses t 1 8 13 5 13 4 82 4 14 125 1 re omitted from the nlysis. n the following nlyses the cses hve been divided into smll reltively homogeneous subgroups nd the tretments compred on stndrdized bsis, though this certinly does not overcome ll the difficulties. The three-yer survivl rtes for cses treted with nd without rdiotherpy re shown seprtely for ech stge in Tble V. The striking difference in survivl between cses receiving preopertive rdiotherpy only nd those receiving preopertive nd postopertive rdiotherpy is probbly explined by the fct tht the ltter group consists lrgely dvnced cses in which extrcpsulr spred hd not been erdicted by the preopertive rdiotherpy. The survivl rte for ll cses treted by preopertive irrdition, combining the groups with nd without postopertive course, is seen to be similr to tht for cses receiving only postopertive rdiotherpy, nd rther worse thn for those receiving none. The cses receiving no rdiotherpy fll into two distinct groups: TBLE V.-Three-yer Survivl Rtes in Wilms's Tumour Relted to Stge nd ge -1 2-3 4- ge t Dignosis (yers) ll ges Stge No. 3-yer Survivors No. Cses No. Cses 54 41 7 7 3 23 5 28 14 5 14 121 78 4 3 _ Cses for which the stging or ge is not vilble re included in "ll Cses." Stge B 3-yer Survivors No., 1 14 3 33 5 3 3 No. Cses 3 47 42 Stge C 3-yer Survivors No. 2 3 3 7 125 8 ll Cses No. 3-yer Survivors Cses No. oi - 11 4 41 115 32 28 1 2 2 335 32 TBLE V.-Three-yer Survivl Rtes in Wilms's Tumour Treted by Nephrectomy with nd without Rdiotherpy Rdiotherpy None.. Preopertive.... With postop. Without postop. Postopertive With chemotherpy Without chemotherpy Stge Stge B Stge C ll Cses No. 3-yer Survivors No. 3-yer Survivors No. 3-yer Survivors No. 3-yer Survivors Survivl Rte Stndrdized for Cses No. Cses No. Cses No. Cses No. Stge nd ge 1 15 7 5 1 2 18 2 11 43 18 42 4 1 8 1 5 12 3 25 7 4 13 33 27 2 33 5 1 2 1 7 3 18 27 12 8 7 7 2 2 1 23 1 43 33 83 52 3 13 5 38 1 214 7 33 35 28 2 7 1 7 3 43 48 5 1 8 3 35 43 55 32 58 2 33 5 1 1 2 123 3 2 2 *Exludes cses which lso hd preopertive rdiotherpy, nd includes cses where chemotherpy dt were not vilble. Cses for which the stging is not vilble re included in "ll Cses."
24 October ctinomycin D in Wilms's Tumour-Ledlie et l. BEC Un MEDCL JOURNL 1 TBLE V.- Three-yer Survivl Rtes for Wilms's Tumour Treted by Nephrectomy with nd without Chemotherpy Chemotherpy ctinomycin Other chemotherpy No chemotherpy.. Stge Stge B Stge C ll Cses No. 3-yer Survivors No. 3-yer Survivors No. 3-yer Survivors No. 3-yer Survivors Survivl Rte _ Stndrdized for Cses No. O Cses No. Cses No. % Cses No. % Stge nd ge --- 1..1..1..l 1 1~~~~~~~~~~~~~/-~~~~~~~~ - ~~~~~~~~~ 1.. 42 2 12 4 7 38 4 2 1 2 2 22 7 47 2 1 Cses for which stging is not vilble re included in "ll Cses." 4 25 5 1 2 1 53 31 2 ntervl to Recurrence fter Nephrectomy TBLE V.- Three-yer Survivl Rtes in Wilms's Tumour for Cses with Recurrence Stge No. 3-yer Survivors Cses No. % 1-5 months.... 15 1 7 1 11-3 months*.... 31 11 35 11 3 27 ll recurrences.. 4 12 2 2 4 2 1 recurrence t 53 months is lso included. Cses for which stging is not vilble re included in "ll Cses." Stge B Stge C No. 3-yer Survivors No. 3-yer Survivors Cses No. % Cses No. % ll Cses Metstses Tretment ncludes Chemotherpy.. +Surgery + rdiotherpy + Rdiotherpy Only chemotherpy.. No chemotherpy +Surgery rdiotherpy + Rdiotherpy Untreted Metstses.. ll cses.. TBLE X.-Three-yer Survivl Rtes in Wilms's Tumour fter Tretment to Metstses Lungs bdomen No. 3-yer Survivors No. 3-yer Survivors Cses Cses No. % No. %. 1-11 18 13 3 23 74 41 52 12 2 14 14 1 4 21 1 2 2 2 22 5 1 12 5 1.. ~~l 25 3 Five cses where dt on tretment or site re not vilble re included in the totls. 15 - - 1 15 53 5 Other & Multiple No. 3-yer Survivors Cses No. % 14 14 35 2 4 2 ~~l~l ~l ~l ~l ~l 2 4 ~l- l- 88 23 35 3 83 28 ll Sites No. 3-yer Survivors Cses No. % 1. ~ ~ 14 8 3 21 1 35 33 12 1 1 stge cses, considered to be such good prognosis tht irrdition ws not required, hd survivl rte 8%; wheres 23 stge B nd C cses, whom 11 died postopertively nd mny the reminder were too ill for further tretment, hd survivl rte only 13 %. The right-hnd column Tble V shows the survivl rtes stndrdized for ge nd stge disese. This llows comprisons between tretments to be mde, tking into ccount the importnce these prognostic fctors. From this it would pper tht the results postopertive rdiotherpy re rther better thn those preopertive, nd tht the ddition chemotherpy improves the survivl rte still further. mong the stge B cses-tht is, those with locl dherence-12 out 3 (4%) received preopertive rdiotherpy compred with 15 % stge nd % stge C cses. The poor prognosis for stge B cses holds both for ptients who hd preopertive rdiotherpy, which probbly cused the dherence, nd for those who did not. The three-yer survivl rtes for ptients receiving nd not receiving chemotherpy s prt their initil tretment re shown in Tble V. Of the 121 cses stted to hve received chemotherpy, ll except nine hd ctinomycin. For stge nd B cses the ptients treted with ctinomycin did rther better thn those not so treted. For stge C cses the difference is much more pronounced: 7 out 54 cses treted by ctinomycin survived to three yers, but only 1 out 5 who received no chemotherpy. For cses receiving ctinomycin the survivl rte ws 38 %, compred with 31 % for those who hd no initil chemotherpy. When these figures re stndrdized to llow for the difference in the types cse treted the figures become 4% nd 2% respectively. ctinomycin ppers to be prticulrly effective in treting the stge C cses. Of the eight stge C cses who survived for three yers or more, seven received ctinomycin nd five postopertive rdiotherpy. t opertion five the eight were found to hve tumour in the inferior ven cv or renl vein, while the other three hd infiltrtion tissues round the kidney. Three the children developed pulmonry metstses: one died more thn three yers lter, but the other two were pprently successfully treted. Possibly some the successes in cses treted with ctinomycin re in fct due to other spects the tretment they received. For exmple, mong stge nd B cses in which recurrence ws observed 14 out 28 (5%) hd surgery for the recurrence, wheres for cses not initilly given chemotherpy the corresponding figure ws only 4 out 4 (%). On the other hnd, the explntion this difference might be tht metstses occurring in ctinomycin-treted cses re more menble to surgicl tretment. gin, the (stndrdized) survivl rte for teching nd children's hospitls is considerbly better thn tht for other hospitls for cses treted with ctinomycin nd rther better for cses not so treted. t my be tht ctinomycin ws used in more effective dosges in the former. Rtes recurrence re consistent with reports tht ctinomycin hs n inhibiting effect on metsttic growth, but our results could esily be ttributed to chnce. Tretment to Metstses Eighteen children in whom metstses were detected fter the initil tretment were still live three yers fter the originl dignosis (three these subsequently died). Three dditionl cses with metstses t the time dignosis hve lso survived. The results shown in Tble V suggest tht recurrences for stge nd B cses hve better prognosis thn those for stge C. This tble lso shows tht the proportion lte recurrences is smllest for stge C, lrger for B, nd lrgest for. This implies tht loclized cses re likely to be t risk recurrence for longer period thn non-loclized ones. Cses with lte recurrences, however, hve better prognosis thn those recurring erly. The results tretment to metstses xe summrized in
2 24 October ctinomycin D in Wilms's Tumour-Ledlie et l. M&DwCLJOURNL Tble X; it is difficult to drw ny conclusions bout the effect individul components tretment. The most effective tretment ppers to hve been surgery. The dditionl benefits rdiotherpy nd chemotherpy re hrd to ssess becuse the smll numbers cses nd possible effects selection. Rdiotherpy lone ws effective, though in smller proportion cses; chemotherpy lone ws never successful. Discussion Recent ppers on the tretment Wilms's tumour hve presented pprently conflicting evidence on the efficcy ctinomycin. Severl ppers hve presented series in which the proportion successes, usully defined s survivl for two yers, is greter for cses treted with ctinomycin thn for cses not given the drug. These rtes rnge from 7 to 2% (Howrd, 15; Frber, 1; Fernbch nd Mrtyn, 1; Burgert nd Glidewell, ). Other uthors quote rtes from 2 to 3 % nd stte tht these re similr to those for their comprison series not given ctinomycin (Sukrochn nd Kiesewetter, 1; Mier nd Hrshw, ; Stone nd Willims, 1). n the present series the overll three-yer survivl rte for cses treted with ctinomycin ws 38%. Three possible explntions for these conflicting reports should be considered: () differences in selection cses, (b) the effects other components the tretment, nd (c) differences in dose regimens. () Not ll ssessments ctinomycin hve mde proper llownce for fctors known prognostic significnce. We feel tht the weight evidence from the ppers cited is tht ctinomycin does improve prognosis, but the informtion vilble in published series is usully indequte for rigorous nlyses to be mde nd vrying degrees cse selection my hve occurred. (b) The effects the other components tretment re even more difficult to nlyse. Most ssessments ctinomycin hve been mde lrgely with cses treted in erlier yers s comprison series, where both initil tretment nd tretment to metstses my hve been different. (c) Burgert nd Glidewell () suggested tht better results re obtined if tretment with ctinomycin is begun t the time opertion rther thn lter, but it is not cler tht the groups cses on which they bse this conclusion re comprble. t hs lso been suggested tht results re improved by giving repeted courses. The best evidence for this is provided by Wolff et l. (18), who found sttisticlly significnt reduction in metstses when compring this type therpy ginst single course in controlled tril. ctinomycin did not become generlly vilble in this country until 15, nd it my be tht its vlue cnnot be properly ssessed until more evidence hs been ccumulted on cses in which dequte dosge regimens hve been used. The present series includes 34 children who received ctinomycin s prt their initil tretment in 1. One child received two courses nd the others one course only. The injections were strted before opertion in nine cses, t opertion in seven, nd postopertively in 15 (for three cses no records re vilble). We conclude tht ech the fctors (), (b), nd (c) my be contributing towrds the different success rtes quoted for ctinomycin nd tht, in prticulr, the reltively disppointing results in the present series could be due to the fct tht the cses were completely unselected nd possibly tht optimum drug schedules were not used. t would pper tht, even dmitting the prcticl difficulties, these questions cn be elucidted only by crrying out controlled tril, in which other drugs, such s vincristine, might lso be included. The evidence in fvour ctinomycin is such tht it seems ethiclly impossible to withhold it t this stge, but we suggest tht further trils similr to tht Wolff et l. (18) should be crried out to evlute vrious different drug schedules, with sufficiently long follow-up to ensure tht recurrences re being prevented nd not simply postponed. The Oxford Survey Childhood Cncers is supported by grnts from the U.S. Public Helth Service (Grnt No. C-532). the Medicl Reserch Council (Grnt No. G.4/23/C), nd the Cncer Cmpign for Reserch. Our thnks re due to the mny doctors nd socil workers, nd in prticulr to the medicl ficers helth nd public helth deprtments nd the records deprtments hospitls, who hve supplied us with the dt. We re grteful to Mrs. J. Ly, the tls Lbortory, Chilton, for ssistnce in progrmming the computer tbultions. Finlly, we cknowledge our indebtedness to the Director the Survey, Dr.. M. Stewrt, for her help in enbling us to ssemble the necessry records nd for her constnt encourgement. REFERENCES Brber, R., nd Spiers, P. (14). Monthly Bulletin the Ministry Helth nd the Public Helth Lbortory Service, 23, 4. Burgert, E. O., nd Glidewell,. (). Journl the mericn Medicl ssocition, 1, 44. Collins, V. P. (155).J'ournl the Louisin Stte Medicl Society,, 474. Collins, V. P., Loeffler, R. K., nd Tivey, H. (15). mericn journl Roentgenology, Rdium Therpy nd Nucler Medicine, 7, 88. Frber, S. (1). Journl the mericn Medicl ssocition, 18, 82. Fernbch, D. J., nd Mrtyn, D. T. (1). Journl the mericn Medicl ssocition, 15, 15. Grci, M., Douglss, C., nd Schlosser, J. V. (13). Rdiology, 8, 574. Howrd, R. (15). rchives Disese in Childhood, 4, 2. Mier, J. G., nd Hrshw, W. G. (). Cncer (Phildelphi), 2,. Miller, R. W. (18). Journl the Ntionl Cncer nstitute, 4,. Person, D., nd Stewrd, J. K. (1). Proceedings the Royl Society Medicine, 2, 85. Pltt, B. B., nd Linden, G. (14). Cncer (Phildelphi),, 1573. Riches, Sir E. W. (editor). (14). Tumours the Kidnev nd Ureter. London, Livingstone. Scott, L. S. (155). British Journl Surgery, 42, 513. Stewrt,., nd Knele, G. W. (). Lncet, 1, 1185. Stone, J., nd Willims,. G. (1). Clinicl Rdiology, 2, 4. Sukrochn, K., nd Kiesewetter, W. B. (1). Journl Peditrics,, 747. Wigger, H. J. (1). mericn Journl Clinicl Pthology, 51, 323. Wolff, J.., Krivit, W., Newton, W.., nd D'ngio, G. J. (18). Neu, Englnd Journl Medicine, 27, 2. Young, D. G., nd Willims, D.. (1). British Journl Hospitl Medicine, 2, 741.